Well, the original proposal that voters passed was SF–LA, which has very large populations at both ends, and makes a bit more sense (this is also what I was counting as 400 miles). The (maybe temporary, maybe not) scale-back seems to be due to rapidly ballooning estimates of construction cost. I agree that a Merced–Bakersfield line is less compelling, to put it mildly.
Which points to what I think is the main issue, both here and with most other US transit projects: US rail construction costs are very high, even compared to other high-wage countries. If we could get subway and rail projects done at the per-mile costs that Japan, Denmark, France, or Germany manage, I think we'd both have more things already successfully built, and more public appetite to fund more.
There's some digging on that question on this blog: https://pedestrianobservations.com/construction-costs/